While almost all monomers containing carbon-carbon double bonds undergo radical polymerization, ionic polymerization is highly selective. This due in part to the stability of the propagating species. Cationic polymerization involves carbenium ions and is essentially limited to those monomers with an electron releasing substituent such as alkoxy, phenyl, vinyl and 1,1-dialkyl; while anionic polymerization involves carbanions and requires monomers possessing electron withdrawing groups such as nitrile, carboxyl, phenyl and vinyl.
Compared to carbanions, which maintain a full octet of valence electrons, carbenium ions are deficient by two electrons and are much less stable and therefore, controlled cationic polymerization requires specialized systems. The instability or high reactivity of the carbenium ions facilitates undesirable side reactions such as bimolecular chain transfer to monomer, β-proton elimination, and carbenium ion rearrangement, all of which limit the control over the cationic polymerization. Typically, low temperatures are necessary to suppress these reactions. Additionally, other considerations such as stabilization of the propagating centers (typically by appropriate choice of counterion and solvent system), use of additives to suppress ion-pair dissociation and undesirable protic initiation, and the use of high-purity reagents to prevent the deactivation of the carbenium by heteroatomic nucleophiles (such as alcohols or amines) are often required. However, if one carefully selects the system, cationic polymerization can display living characteristics.
Through these living cationic systems, cationic polymerization can be controlled to yield tailored polymers with narrow molecular weight distributions and precisely controlled molecular weight, micro-architecture, and end group functionality. Controlled cationic polymerizations are deemed to be achieved under conditions in which chain end termination is reversible (quasiliving conditions) and undesirable reactions such as chain transfer and water-initiation are suppressed. A tremendous advantage of living and quasiliving polymerization is the opportunity for one-pot in situ functionalization of the polymer by reaction of the living chain ends with an appropriate quenching reagent. Historically, commercial functionalization of oil and fuel additive polymers has been a complex multi-step process. However, commercial implementation of in situ functionalization could reduce the time, energy, and overall cost associated with the production of oil and fuel additives. For example, polyisobutene-based oil dispersants are typically produced by first polymerizing isobutene (IB) to form an olefin-terminated polyisobutene (PIB), reacting the PIB with maleic anhydride to form PIB-succinic anhydride (PIBSA), and then reacting PIBSA with a polyamine to form a PIB-succinimide amine. In total, the dispersant requires three synthetics steps; each stage requires separate reaction conditions and exhibits less than 100% yield.
Living polymerizations refer to any polymerization during which propagation proceeds with the exclusion of termination and chain transfer and thus yields polymers retaining (virtually indefinitely) their ability to add further monomer whenever it is supplied to the system. This description is often too rigorous for actual systems and is approximated herein by quasiliving carbocationic polymerization (QLCCP), which includes chain growth polymerizations that proceed in the absence of irreversible chain breaking mechanisms during the effective lifetime of monomer consumption.
With the advent of carbocationic living polymerization and QLCCP, there have been attempts to functionalize these living polymers. The extent of success of these attempts has been directly linked to the type of monomer being polymerized. Simple one pot (or in situ) chain end functionalization of more reactive carbocationic monomers, like isobutyl vinyl ether, can occur using ionic nucleophilic quenching reagents, i.e. methanol, alkyl lithium etc., Sawamoto, et al. Macromolecules, 20, 1, (1987). However chain end functionalization does not occur when these reagents are added to living polymerization of less reactive monomers such as isobutylene, Fodor et al. Polym Prepr. Amer. Chem. Soc., 35(2), 492 (1994). Addition of these reagents at the end of polymerization resulted in the consumption of the catalyst and the formation of tert-alkyl chloride chain ends on the polyisobutylene (PIB) rather than the desired nucleophilic substitution. This represented a trivial result since QLCCP of IB produces tert-chloride end groups anyway, as a direct consequence of the inherent, reversible termination mechanism in these polymerization systems.
Functionalization of quasiliving PIB has typically been attempted through the use of functional initiators and through in situ functionalization by quenching. Most past efforts to produce functionality by quenching of quasiliving chains has failed and has led to the tert-chloride terminus, J. Macromol. Sci-Chem, A27, 649 (1990) and Polym. Bull, 18, 123 (1987). The accepted rationale is that quasiliving PIB is composed primarily of dormant (reversibly terminated) chains. Thus, most added reagents, particularly strong nucleophiles, quench the Lewis acid co-initiator and therefore yield only the tert-chloride chain end. Tert-chloride groups are often undesirable as a dispersant/detergent for lubricants and fuels due to environmental reasons and since their presence may decrease the effectiveness of controlling soot and other engine contaminants. Additionally, tert-chloride groups tend to decompose, liberating HCl, which is corrosive toward metal surfaces within the engines. The most notable exception to the above general rule was the discovery that allyltrimethylsilane (ATMS), when added in excess to living polyisobutylene, does not react with the Lewis acid but rather is alkylated by the PIB chain end, thereby providing living PIB with allylic ends groups in situ, U.S. Pat. No. 4,758,631. A related U.S. Pat. No. 5,580,935 teaches the use of alkylsilylpseudohalides as quenching agents, thereby adding to the choice of chemistries. However, functionalization of the cationic polymers in situ with suitable nitrogen compounds for use in dispersants and/or detergents has been elusive.
Based upon the success of ATMS, Faust et al. investigated 2-substituted furan derivatives and found that quantitative reaction with quasiliving PIB chain ends could be achieved in both titanium tetrachloride (TiCl4) and BCl3 co-initiated systems; Macromolecules 32, 6393 (1999) and J Macromol., Sci Pure Appl. Chem. A37, 1333, (2000). Similarly, Ivan in WO 99/09074 disclosed quenching quasiliving PIB with furan derivatives and thiophene derivatives while postulating that any aromatic ring, preferably 5-7 membered heterocycles as well as optionally substituted moieties could be employed to quench and effectively functionalize QLCP PIB through electrophilic aromatic substitution. We have now found that there is particularity of the aromatic ring, the substituent group on the ring, as well as the position of the substituent group on the ring. Incorrect selection of the aromatic ring or substituent, such as substituents which contain certain nucleophile segments (such as —OH, —NH2) can deactivate the catalyst and render the PIB chain end unaffected and carrying only tert-Cl end groups, or in certain circumstances, couple the quasiliving polymer. The present invention is based partly on the discovery that monodisperse telechelic polymers can be produced by cationic polymerization of a suitable monomer under living polymerization conditions and that quenching the polymerization with an N-substituted pyrrole yields PIB with desirable nitrogen-containing end functionality. Such N-substituted pyrroles are unique since they contain a tertiary nitrogen atom with the lone pair of electrons taking part in the aromatic sextet of electrons in the 5-membered, aromatic pyrrole ring and surprisingly, produce high yields of monodispersed chain end functionalized polymer.